Physical activity declines during adolescence, particularly among girls, and this decline substantially contributes to the current child/adolescent obesity epidemic. Physical activity at a given point in time is a highly complex behavior, and the dynamic process of decline is inherently more complex. What's more, because of the rapid physical, social, and emotional growth that occurs during adolescence, factors associated with physical activity decline are likely to change from early to late adolescence. Multi-level ecologic models that include the context in which people live (i.e., policy, environment, social, individual) are now considered state-of-the-art for describing health behaviors, including physical activity. Yet these multi-level ecologic frameworks have not been used to examine physical activity in adolescent girls, particularly the factors influencing the decline at different phases of adolescence. Our investigative team is uniquely positioned to examine the complexity of physical activity decline in girls across adolescence. Through our participation in TAAG, we have policy, environmental, social, and individual-level data on two large cohorts of girls measured during early adolescence: 1) a nationally representative cohort of 984 girls who participated in TAAG (followed from 6th to 8th grade), and 2) 700 racial/ethnically diverse 8th grade girls residing in Maryland. In this application we propose to conduct a longitudinal follow-up of these girls who are now in mid-late adolescence (i.e., 11th grade;"Physical Activity High School Decline" [PAHSD];new data). Our goal is to identify social, environmental, and individual-level factors that contribute to the decline in physical activity during early and mid-late adolescence. We propose to address the following specific aims: (1) To examine neighborhood, school, social, and individual-level factors associated with the decline in physical activity in girls in early adolescence (6th to 8th grade, TAAG) and mid-late adolescence (8th to 11th grade, PAHSD), and (2) To determine how factors associated with the decline in physical activity differ in early adolescence and mid-late adolescence. Given the national concern about the increasing prevalence of adolescent obesity, we will also examine how neighborhood, school, social, and individual-level factors are associated with change in body composition in early to mid-late adolescence. PUBLIC HEALTH RELEVANCE: Physical inactivity contributes to poor health among adolescents, and influences their future health as adults. To halt the physical activity decline in adolescents, we must understand factors associated with the decline. This proposed study will examine multi-level predictors of physical activity decline in adolescent girls, which will provide information for future intervention design to halt the decline.